Tiwe
|
  |
| Joined: 25 Aug 2008 |
| Total Posts: 18450 |
|
|
| 23 Oct 2013 04:34 AM |
Every pixel on a shirt is considered significant and taken into account so the idea behind this is to take note of the the numbers of pixels on a shirt then split them into tables with their respective color data. The percentage of colored pixels on a shirt would then be displayed.
Shirts follow the same template so they will always have the same number of pixels on them, if you count transparent or blank ones which you should. The system then counts how many pixels and of what color each of them are. Group them together by color and then get the percentage of each color used on the shirt (including transparency).
Every time a shirt is uploaded, the database compares the data gathered and if the numbers (not the relying on the colors since it can easily be changed) on the percentages of the shirt match/have slight the another shirt then it's up for "high priority" and because moderation isn't exactly a strong point here and since it isn't 24/7, shirts in high priority will be delayed even further and the users will be notified of the delay.
So here's how it goes. Player A makes a shirt and uploads it. Database notes its pixel composition; let's say 50% red, 7% black, 4% transparent, and 39% orange. The database compares the gathered data with other shirts and it can find either one of the three:
a. Complete Match (same colors, same percentages), high priority (not an automated reject since there's still a small chance that the shirts aren't identical but just composed in a similar manner).
b. Slight Difference (different colors but same percentages, the same colors but a minor difference in percentages, different colors but a minor difference in percentages), the minor differences are up to the margin of error and it's of high priority.
c. Unique and passable
- One idea to strengthen this system is by taking note of pixel positions and it just goes through the same process. It's still going to be flawed. This has many flaws. It's an idea and it's an idea with the purpose of sparking up what others are thinking of regarding this matter in their head right now.
This isn't really a suggestion asking for a support, it's more of a suggestion to let other people read and then leave off to make their own suggestions. ~ There is no implication on the level of coding here.
|
|
|
| Report Abuse |
|
|
Tiwe
|
  |
| Joined: 25 Aug 2008 |
| Total Posts: 18450 |
|
|
| 23 Oct 2013 10:33 AM |
| Final bump of the lifetime. |
|
|
| Report Abuse |
|
|
|
| 23 Oct 2013 10:37 AM |
This concept could probably be exploited, but its definitely a start.
Support. |
|
|
| Report Abuse |
|
|
BIOWARIOR
|
  |
| Joined: 04 Apr 2010 |
| Total Posts: 506 |
|
| |
|
|
| 23 Oct 2013 11:15 AM |
This would just increase CPU usage and decrease database space. I do not support this suggestion. |
|
|
| Report Abuse |
|
|
fanGLaDOS
|
  |
| Joined: 25 Apr 2012 |
| Total Posts: 89 |
|
|
| 23 Oct 2013 12:40 PM |
| No, this idea is great! I don't want my shirt copied - If I made one. |
|
|
| Report Abuse |
|
|
Tiwe
|
  |
| Joined: 25 Aug 2008 |
| Total Posts: 18450 |
|
|
| 07 Nov 2013 06:45 AM |
| Here's a bump, just to hope that people can take something from reading this and suggest a much better planned and thought out idea. |
|
|
| Report Abuse |
|
|
Z007
|
  |
| Joined: 14 Sep 2010 |
| Total Posts: 11729 |
|
|
| 07 Nov 2013 07:27 AM |
That is actually a really good point, Animated.
Let's take into account how slow your computer will become when you upload a shirt. Frankly, I'm not sure how fast their databases are, but it's probably not the fastest when you're searching EVERY SINGLE SHIRT. EVERY ONE.
After that, you've got cost. I highly doubt this'd be budget-friendly in any way, shape, or form. As such, prices in the catalog will go up. |
|
|
| Report Abuse |
|
|